Age-related macular degeneration (AMD) is a leading cause of blindness in the elderly population of Western countries. In the past few years, over one dozen AMD risk loci have been identified through genome-wide association studies (GWAS), either by individual studies or through meta-analyses of multiple studies from the National Eye Institute (NEI) supported AMD Gene Consortium. An ongoing exome chip experiment on 38,000 AMD/Control subjects will further expand the list by discovering additional rare variants. However, the analyses and statistical methods are still lagging behind the pace of data generation. Emerging genetic and phenotypic data from our collaborators, the AMD Exome Chip Consortium, and public databases (e.g. the dbGaP) will allow us to test new hypotheses, develop and calibrate statistical methods to facilitate ongoing consortium studies in which we are involved. In particular, we are interested in systematically studying the genetic causes and prediction of AMD progression, identifying disease-susceptibility loci in a cohort of African Americans, and developing association methods for family-based studies with binary traits. To achieve these goals, we propose specific aims as follows: 1) To develop a bivariate survival framework to jointly model AMD progression in both eyes and to perform a genome-wide association study of AMD progression using over 4,000 eligible samples from AREDS (Age-Related Eye Disease Study), AREDS2, and the AMD study conducted at the University of Michigan; 2) To develop and validate rigorous statistical models for prediction of AMD occurrence and progression based on demographic, clinical, and genetic information from the results of Aim 1 and to obtain predictive probabilities accounting for different study designs and the correlation between two eyes; 3) To develop and apply novel methods to identify loci associated with AMD risk in 725 unrelated African Americans, combining signals from both association and admixture mapping; and 4) To develop a statistical method for rare variant association tests of binary traits in families under the framework of generalized linear mixed model using a functional modeling approach and to apply the method to our UCLA- Pittsburgh family-based study of 2,188 samples. Our results will advance our understanding of pathogenesis and prevention of AMD occurrence and its progression. The methods we developed and applied will be available to other study groups and will benefit the analysis of ongoing AMD consortium data. In addition, our methods can be applied to other vision research as well. Unique strengths of our research team include: extensive prior experience in the applied analyses of AMD data sets, outstanding statistical genetics expertise, and clinical consultants with deep insight into the AMD data sets they collected. Successful completion of our Aims, where we will develop and apply state-of-the-art statistical methods, will enrich our understanding of AMD pathogenesis and improve individual risk prediction, and therefore will help enhance clinical practice.